f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001

[cecilehannay]

Description:
FMTHIST model run with the same tuning as #282 and:

• effgw_beres_dp = 0.7
• add daily TEM diagnostics
• add HB (Holtslag and Boville) boundary layer scheme where CLUBB is not active.
• add aircraft emissions

user_nl_cam

Tuning/emissions for this run:

clubb_l_predict_upwp_vpwp=.true. 
clubb_l_mono_flux_lim_um   = .true.  
clubb_l_mono_flux_lim_vm   = .true.  
clubb_c_uu_shr = 0.1
clubb_c7=0.1

dust_emis_fact         = 0.80D0

micro_mg_dcs=0.00025
micro_mg_vtrmi_factor=0.5
microp_aero_wsubi_scale=2.5
seasalt_emis_scale=1.5

use_gw_convect_dp = .true.
use_gw_front =.true.
tau_0_ubc = .true.

effgw_beres_dp = 0.7 

effgw_rdg_beta = 0.5D0
effgw_rdg_beta_max = 0.5D0


 ext_frc_specifier		= 'num_a1 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_num_so4_a1_anthro-ene_vertical_mol_175001-201412_ne30pg3_c20200103.nc',
         'num_a1 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_num_a1_so4_contvolcano_vertical_850-5000_ne30pg3_c20200125.nc',
         'num_a2 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_num_a2_so4_contvolcano_vertical_850-5000_ne30pg3_c20200125.nc',
         'SO2 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_SO2_contvolcano_vertical_850-5000_ne30pg3_c20200125.nc',
         'so4_a1 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_so4_a1_anthro-ene_vertical_mol_175001-201412_ne30pg3_c20200103.nc',
         'so4_a1 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_so4_a1_contvolcano_vertical_850-5000_ne30pg3_c20200125.nc',
         'so4_a2 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_so4_a2_contvolcano_vertical_850-5000_ne30pg3_c20200125.nc'
	 'bc_a4 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_bc_a4_air_vertical_175001-201412_ne30pg3_c20200521.nc'
	 'num_a4 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_num_bc_a4_air_vertical_175001-201412_ne30pg3_c20200521.nc'
	 'SO2 -> /glade/p/cesmdata/cseg/inputdata/atm/cam/chem/emis/historical_ne30pg3/emissions-cmip6_SO2_air_vertical_175001-201412_ne30pg3_c20200522.nc'

Output will be:

mfilt		    =       0,       5,     20,      40,      12,       120,      1,   1
nhtfrq              =       0,     -24,    -24,      -3,       0,       -2,      0,  -8760
ndens               =       2,       2,      2,       2,       2,       1,      2,   1
interpolate_output  =  .true.,  .true., .true., .false., .false., .true.,  .true.
interpolate_nlat    =     192,     192,    192,     192,     192,     192,   192
interpolate_nlon    =     288,     288,    288,     288,     288,     288,   288 

empty_htapes = .true.

fincl1 = 'ACTNI', 'ACTNL', 'ACTREI', 'ACTREL', 'AODDUST', 'AODVIS', 'AODVISdn','BURDENBC', 'BURDENDUST', 'BURDENPOM', 'BURDENSEASALT', 
'BURDENSO4', 'BURDENSOA', 'CAPE', 'CCN3', 'CDNUMC', 'CH4', 'CLDHGH', 'CLDICE', 'CLDLIQ', 'CLDLOW', 'CLDMED', 'CLDTOT', 'CLOUD', 'CMFMC_DP', 
'CT_H2O', 'DCQ', 'DQCORE', 'DTCOND', 'DTCORE', 'DTV', 'EVAPPREC', 'EVAPSNOW', 'FCTI', 'FCTL', 'FICE', 'FLDS', 'FLNS', 'FLNSC', 'FLNT', 'FLNTC', 'FLUT', 
'FREQZM', 'FSDS', 'FSDSC', 'FSNS', 'FSNSC', 'FSNT', 'FSNTC', 'FSNTOA', 'ICEFRAC', 'LANDFRAC', 'LHFLX', 'LWCF', 'MPDICE', 'MPDLIQ', 'MPDQ', 'MPDT', 
'OCNFRAC', 'OMEGA', 'OMEGA500', 'PBLH', 'PHIS', 'PINT', 'PMID', 'PRECC', 'PRECL', 'PRECSC', 'PRECSL', 'PRECT', 'PS', 'PSL', 'PTEQ', 'PTTEND', 'Q', 
'QFLX', 'QRL', 'QRS', 'QTGW', 'RCMTEND_CLUBB', 'RELHUM', 'RVMTEND_CLUBB', 'SHFLX', 'SOLIN', 'SST', 'STEND_CLUBB', 'SWCF', 
'T', 'TAUX', 'TAUY', 'TFIX', 'TGCLDIWP', 'TGCLDLWP', 'TMQ', 'TREFHT', 'TS', 'TTGW', 'U', 'U10', 'UBOT', 'UTGWORO', 'UTGW_TOTAL', 
'V', 'VBOT', 'VTGWORO', 'VTGW_TOTAL', 'WPRTP_CLUBB', 'WPTHLP_CLUBB', 'Z3', 'ZMDQ', 'ZMDT', 'N2O', 'CO2','CFC11','CFC12', 
'CLD_MISR','FISCCP1_COSP','CLD_CAL','CLD_MISR','CLDTOT_CAL','CLDHGH_CAL', 'CLDMED_CAL','CLDLOW_CAL','CLMODIS', 'AODVISdn',
'CCN3', 'CDNUMC','OMEGA500'


fincl2 = 'PS','PMID','U','V','T','Q'

fincl3 = 'PRECT', 'PRECC', 'FLUT', 'U850', 'U200', 'V850', 'V200', 'OMEGA', 'PSL',  'TS','PS'

fincl4 =  'PRECC','PRECL'

fincl5 = 'Uzm','Vzm','Wzm','THzm', 'VTHzm','WTHzm','UVzm','UWzm'

fincl6 = 'Uzm','Vzm','Wzm','THzm', 'VTHzm','WTHzm','UVzm','UWzm'

phys_grid_ctem_nfreq=-6
phys_grid_ctem_zm_nbas=120
phys_grid_ctem_za_nlat=90

fincl7= 'AQSO4_H2O2','AQSO4_O3', 'bc_a1', 'bc_a4', 'dst_a1', 'dst_a2', 'dst_a3', 'ncl_a1',
'ncl_a1', 'ncl_a2', 'ncl_a3', 'pom_a1', 'pom_a4', 'so4_a1', 'so4_a2', 'so4_a3',
'soa_a1', 'num_a1', 'num_a2', 'num_a3', 'num_a4',
'bc_a1SFWET', 'bc_a4SFWET', 'dst_a1SFWET', 'dst_a2SFWET', 'dst_a3SFWET', 'ncl_a1SFWET',
'ncl_a2SFWET', 'ncl_a3SFWET', 'pom_a1SFWET', 'pom_a4SFWET', 'so4_a1SFWET', 'so4_a2SFWET', 'so4_a3SFWET', 'soa_a1SFWET',
'soa_a2SFWET', 'bc_c1SFWET', 'bc_c4SFWET', 'dst_c1SFWET', 'dst_c2SFWET', 'dst_c3SFWET', 'ncl_c1SFWET', 'ncl_c2SFWET',
'ncl_c3SFWET', 'pom_c1SFWET', 'pom_c4SFWET', 'so4_c1SFWET', 'so4_c2SFWET', 'so4_c3SFWET', 'soa_c1SFWET', 'soa_c2SFWET',
'bc_a1DDF', 'bc_a4DDF', 'dst_a1DDF', 'dst_a2DDF', 'dst_a3DDF', 'ncl_a1DDF', 'ncl_a2DDF', 'ncl_a3DDF',
'pom_a1DDF', 'pom_a4DDF', 'so4_a1DDF', 'so4_a2DDF', 'so4_a3DDF', 'soa_a1DDF', 'soa_a2DDF',
'so4_a1_CLXF', 'so4_a2_CLXF', 'SFbc_a4', 'SFpom_a4', 'SFso4_a1', 'SFso4_a2',
'so4_a1_sfgaex1', 'so4_a2_sfgaex1', 'so4_a3_sfgaex1', 'soa_a1_sfgaex1', 'soa_a2_sfgaex1',
'SFdst_a1','SFdst_a2', 'SFdst_a3', 'SFncl_a1', 'SFncl_a2', 'SFncl_a3',
'num_a2_sfnnuc1', 'SFSO2', 'OCN_FLUX_DMS', 'SAD_SULFC', 'SAD_TROP', 'SAD_AERO'

user_nl_clm

use_init_interp = .true.
check_finidat_year_consistency = .false.
fsurdat = '/glade/work/slevis/git/mksurfdata_toolchain/tools/mksurfdata_esmf/surfdata_ne30np4.pg3_SSP5-8.5_78pfts_CMIP6_1850-2100_c230227.nc'
flanduse_timeseries = '/glade/work/slevis/git/mksurfdata_toolchain/tools/mksurfdata_esmf/landuse.timeseries_ne30np4.pg3_SSP5-8.5_78_CMIP6_1850-2100_c230227.nc'

Case directory:
Locally (if still available):
/glade/p/cesmdata/cseg/runs/cesm2_0/f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001

On github:
https://github.com/NCAR/amwg_dev/tree/f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001

Source Mods
HB scheme:

/glade/u/home/pel/src/for-hb-nohack-cam6_3_110/vertical_diffusion.F90
/glade/u/home/pel/src/for-hb-nohack-cam6_3_110/clubb_intr.F90

CLM mods

/glade/u/home/oleson/run_hist_1850_files/SPINUP/casefiles_cecile/SourceMods/ctsm5.1.dev120/src.clm/

Sandbox:
Locally (if still available):
/glade/work/hannay/cesm_tags/cam6_3_112

On github:
(https://github.com/ESCOMP/CAM/tree/cam6_3_112)

Diagnostics:
ADF diags (if available)
https://webext.cgd.ucar.edu/FMTHIST/f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001/atm/

Contacts:
@tilmes

[justin-richling]

ADF vs f.cam6_3_112.FLTHIST_v0c.ne30.non-ogw-ubcT-effgw0.3.001 here

ADF vs Obs here

[justin-richling]

Updated ADF vs Obs here

[dan800]

@rgarc1a @islasimpson @JulioTBacmeister
I had a quick look at the H2O in the model. I am not seeing anything odd at the upper boundary with a zero flux condition.
There looks to be too much dehydration at the high latitude in SH stratosphere. These zonal means are from the end of the run and compared to WACCM.

FMTHIST

WACCM

FMTHIST

WACCM

Time series at the lid:

The methane looks pretty good:
FMTHIST

WACCM

[rgarc1a]

"I had a quick look at the H2O in the model. I am not seeing anything odd at the upper boundary with a zero flux condition. These zonal means are from the end of the run and compared to WACCM" You really don’t see any difference between, for example, these two NH summer distributions? R.

On Jun 9, 2023, at 4:04 PM, Dan Marsh ***@***.***> wrote:

Rolando Garcia NCAR/ACOM P.O. Box 3000 Boulder, CO 80307-3000 (303) 497-1446 ***@***.***

[dan800]

You really don’t see any difference between, for example, these two NH summer distributions? R.

I said I don't see anything odd at the UB, not that they are not different. i.e., there are no concerns about a build up of H2O at the upper boundary. In the stratosphere, outside the region of excessive dehydration, the vmrs look to match to within 1 ppmv, which is probably better than prior versions of CAM. If we can get an update on the loss rates for CH4 from the 30 year old 2-D loss rates and can tune the entry value into the stratosphere we might even get this closer. Now, that excessive dehydration is worrying - can that just been down to a larger cold-pole problem in MT CAM vs WACCM or is there an issue with how dehydration is being handled?

I'd add that in this model, I think the main concern is to get the H2O close in the stratosphere where it contributes 5 or 10% of the total LW cooling. In the mesosphere, it's not a big player, so errors up there are not that detrimental to the radiative budget (and it's effectively in the sponge anyway).

[rgarc1a]

Specifically about water vapor in NH summer (3rd and 4th panels above): There is extreme dehydration over Antarctica, but that is probably due to too low T there, which presumably could be fixed. But the H2O vmr and gradients above 10 hPa are not realistic in MT and they may get worse with time. The tape recorder also looks different, although perhaps this run has not progressed long enough to tell.

What does total hydrogen look like? (H2O + 2*CH4)

[dan800]

Specifically about water vapor in NH summer (3rd and 4th panels above): There is extreme dehydration over Antarctica, but that is probably due to too low T there, which presumably could be fixed. But the H2O vmr and gradients above 10 hPa are not realistic in MT and they may get worse with time. The tape recorder also looks different, although perhaps this run has not progressed long enough to tell.

This is at the end of the run, and there doesn't seem to be any strong drifts. A more quantitative comparison:

At 10 hPa the gradients look pretty similar outside of that SH polar bias - MT looks to be 0.5 ppmv wetter.
At 1 hPa the diffs are still about 0.5 ppmv but theres a difference in gradient in the NH. However the polar bias is very much worse. As above, not sure this is coming all from a T bias.

[rgarc1a]

More on total hydrogen: As I recall, in WACCM, H2O + 2CH4 is pretty much constant through 0.1 hPa and begins to decline thereafter, and H2 increases. So at 0.01 hPa you may want to check H2O + 2CH4 + H2. Eventually even H2 is broken down and there is a flux of H through the upper boundary (loss to space), which in WACCM is implicitly imposed by using H vmr from MSIS as the UBC.

Loss of H to space may not matter if it is small enough on the time scales of interest, which I presume is what you are saying. In fact, you may effectively be removing more H2 (as H2O) from the stratosphere in MT than in WACCM due to the extreme dehydration over Antarctica. It would be interesting to compare H2O + 2*CH4 + H2 in the two runs.

If all one wants to do is tropospheric climate, none of this may be important as long as one gets tropical H2O about right. Is that your point?

[tilmes]

@rgarc1a Remember we want to use the MT model run for prescribing chemistry and oxidants, this is not just for the tropospheric climate, but we need to get the stratosphere also as good as we can.

[rgarc1a]

Re 10 and 1 hPa profiles of H2O. The values are clearly different. Otherwise, yes, the profiles ought to look similar, especially if you tau0_ubc = true.

[dan800]

At 10 hPa they match to about 10%. This is the area where I would hope we would want to get the LW cooling rates to be as best as we can. As a GHG, water above 1 hPa is not a major player at all, so the only reason to get it close is to ensure water transported into the stratosphere by the residual circulation doesn't create issues. The bigger fish to fry is that massive dehydration!

Looking at the tropics lower in the atmosphere indicates this 0.5 ppmv wet bias at 10 hPa is coming from differences in the entry value (see updated time series at 72 hPa below - WACCM has been converted to mmr). Before we had too dry a value, now we seem to have a value that is too moist. @adamrher may want to comment on what caused that. Doug has the definitive dataset to compare to.

Note @tilmes, that what we see here may not carry over to runs with chemistry - these runs are using loss rates from the NOAA (G-S 2D) model. Doug says he will work on updating them.

[rgarc1a]

197 hPa is not in the stratosphere in the Tropics. From the above plot it looks like the tropical upper troposphere is wetter in MT.

[rgarc1a]

Notice, BTW, there is no tape recorder in the above plot. That's because the water has not experienced the cold point yet. I presume this is a specified SST run (AMIP), judging from the big wad of H2O in 1998.

[adamrher]

Before we had too dry a value, now we seem to have a value that is too moist. @adamrher may want to comment on what caused that.

We did see that the new F tunings moisten the UTLS. I have not checked to see what the HB mods do to moisture. If I wasn't working of workshop talks, I'd engage in the Friday night analysis w/ y'all.

[tilmes]

@JulioTBacmeister @islasimpson @dan800 @rgarc1a
Please check out the old AMWG diagnostics that include the WACCM comparisons:
The MT run is compared to a FV WACCM 1deg nudged run we did a while ago. The temperatures and winds are from MERRA2 so they should be a good reference.
Left is the MT run, right is the old WACCM.
https://acomstaff.acom.ucar.edu/tilmes/amwg/cam7/f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001.1996_2004-c_cesm2.1b01_fswd_1975_cntrl.1996_2004/

Here are a few things to point out, it would be great to do this before running chemistry (my opinion) but if you all think we should run with these settings with Chemistry, let me know.
Please send any suggestions for new tuning settings.

1. SH polar temperatures SON: still very cold, there is a very strong positive bias at the top of the model, not sure if this is something we can change?

2. SH jets SON: still way too strong

3. Water vapor: too low at the model top, too high above the tropopause

... You can check out the other plots..

[tilmes]

I can setup a shorter Chemistry run to see if some of the issues are related to the GHGs reactions, as we have them currently? This will however not fix the cold vortex, I assume.

[dan800]

Before we had too dry a value, now we seem to have a value that is too moist. @adamrher may want to comment on what caused that.

We did see that the new F tunings moisten the UTLS. I have not checked to see what the HB mods do to moisture. If I wasn't working of workshop talks, I'd engage in the Friday night analysis w/ y'all.

Before we engage in tuning to try to match WACCM, we should see which is actually producing an entry value close to MLS (or whichever is the best obs. dataset).

[dan800]

Please check out the old AMWG diagnostics that include the WACCM comparisons: The MT run is compared to a FV WACCM 1deg nudged run we did a while ago. The temperatures and winds are from MERRA2 so they should be a good reference. Left is the MT run, right is the old WACCM.

Thanks, Simone!

1. SH polar temperatures SON: still very cold

Yes, this is a major concern, especially when running chemistry.

there is a very strong positive bias at the top of the model, not sure if this is something we can change?

The bias at the top could be from not running non-LTE cooling, which in WACCM comes into effect starting at 65 km. It's an inexpensive parameterization and we might explore turning it on. At which level is MERRA2 nudging turned off in this run - is this a WACCM - MT bias or a MERRA2 - MT bias?

3. Water vapor: too low at the model top, too high above the tropopause

Yes, as shown above - there's a 0.5 ppmv entry value difference.

[rgarc1a]

Do we even know what WACCM SE does with water in the UTLS? I see that the former 197 hPa H2O time series was replaced with one for 72 hPa, which does show TR. Is the WACCM case from WACCM SE or is this an old FV run?

[dan800]

Is the WACCM case from WACCM SE or is this an old FV run?

The WACCM case is an ensemble of FV runs "f.e21.FWHISTBgcCrop.f09_f09_mg17.CMIP6-AMIP-WACCM.ensAvg123." But that doesn't mean it's right - we need to wait for Doug to comment as to the observed seasonal cycle in the lower tropical stratosphere. Simones comparison is with a different nudged run, but shows a similar bias.

[rgarc1a]

"The WACCM case is an ensemble FV run "f.e21.FWHISTBgcCrop.f09_f09_mg17.CMIP6-AMIP-WACCM.ensAvg123." But that doesn't mean it's right"

Yes, of course. But the point was that we are now in the SE era, so comparisons with WACCM should be against SE WACCM. As regards whether FV WACCM was "right", you might recall that going to 500 m vertical resolution (in the free troposphere and lower stratosphere) dehydrated the upper troposphere and the tape recorder (particularly the wet phase) relative to standard 70L WACCM. Standard WACCM agreed better with MLS H2O, but then the latest revision of MLS data lowered the vmr somewhat, so 110L WACCM (500 m resolution) did not look so bad anymore.

Clearly, careful examination of MLS versus SE WACCM (500-m resolution) is needed. In the meantime, if one needs to evaluate MT, one can do that directly against MLS.

[tilmes]

@dan800 @rgarc1a @dkinnison
Here are comparisons to observations:
https://acomstaff.acom.ucar.edu/tilmes/amwg/cam7/f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001-obs.1996_2004/

[rgarc1a]

Simone, I took a quick look at "f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001"
water vapor (85 hPA) in the above comparison (see below). The run is labeled "f.cam6_3_112.FMTHIST_v0c.ne30.non-ogw-ubcT-effgw0.7.001". I have no idea whether this is LT or MT. It does look like a SE run ("ne30"). Whatever it is, the TR looks way too wet compared to MLS. And I am not sure what version of MLS water vapor that is, either. (The diagnostics web site indicated that the MLS data goes through 2016, so I think this is not the latest update.)

I think a lot more clarity is required to make sense of any of this.